Re: [dns-wg] DNS registry sofwtare (Was: Ripe 46 Meeting

[Edward Lewis <]

I highly encourage folks who are using any IETF reviewed document to 
be sure to read and understand this document before making a 
judgement on the status of any development within the IETF:

  RFC 2026: http://ietf.org/rfc/rfc2026.txt

As far as EPP is concerned, Jaap's words are correct (as would be 
expected of a co-chair).  It is true that earlier versions of EPP are 
running around, but keep in mind that EPP as described in the 
documents in the RFC editor queue have these advantages over the 
earlier versions:

1) The RFC Editor-Queued documents reflect more and more public 
review of the work, hence a more broadly acceptable proposal than 
earlier documents.  This is of course relative to those who actively 
participated in the effort.

2) The Queued documents have been reviewed by the IESG process 
representing a broader range of expertise to make sure EPP is more in 
line with the internet protocols previously defined.

3) The Queued documents will be archived "permanently" unlike the 
documents that describe the earlier versions.

The currently Queued documents represent a Proposed Standard in IETF 
language, which promises higher quality than an internet draft, but 
lacks sufficient operational testing to be considered a Standard fit 
for full blown production.  On one hand, the Queued documents 
represent a stable and qualified document definition.  On the other 
hand, there may be unknown issues that will need tweaking before we 
are satisfied that EPP is solid.

There's no (guaranteed) free lunch.  The IETF consists of folks 
volunteering (perhaps because their employer pays them to do so) to 
make a protocol definition.  The IETF does not provide free software 
(although participants may do so), nor sell software (ditto), nor 
does it require the use of any protocol or software in the network. 
The IETF is just a group of engineers trying to make protocols 
possible.  Operators (registries too) will need invest to benefit 
from the IETF definitions.

At 12:08 +0200 7/17/03, Stephane Bortzmeyer wrote:
> Most big EPP registries do *not* use the current version (the one in
> the RFC Editor queue).
>
> Most free EPP software do *not* use the current version (and hence do
> not interoperate, for instance Gandi's IRI, which works with ".biz"
> and ".info" registries does not work with OpenReg).
Remember that "you get what you pay for" - I'm not saying that the 
free software is of poor quality - but if the software base isn't 
conformant to the Queued documents (better yet the resulting RFCs), 
I'd be wary to count on it for my operations.

--
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Edward Lewis                                            +1-703-227-9854
ARIN Research Engineer

...as graceful as a blindfolded bull in a china shop...